Lewis Blake

Computational Scientist

Curriculum Vitae

Lewis Blake

Computational Scientist

 lewis.blake@met.no

Education

Experience

  • Computational Scientist, Norwegian Meterological Institute, March 2022 - Present

    • Researcher and software developer in the Department of Research and Development, Division For Climate Modeling and Air Pollution
    • Managed and implemented contracts for the Copernicus Atmosphere Monitoring Service (CAMS) for regional and global air quality model evaluation from software package development through production
    • Achieved significant advancements in pollution deposition modeling for Norway, utilizing cutting-edge machine learning-based data fusion techniques
    • Authored a comprehensive technical report outlining the methodology, findings, and environmental and economic impacts for the Norwegian Environment Agency
  • Graduate Research Assistant, NCAR - National Center for Atmospheric Research, Summers 2018, 2019, and 2021
    • Contributed to the Computational and Informational Systems Laboratory, Analytics and Integrative Machine Learning Group during summer appointments in 2018, 2019, and 2021
    • Developed and implemented highly efficient parallel machine learning models (Multi-Resolution Approximation for Gaussian Processes), enabling processing of large datasets (hundreds-of-millions of observations) with reduced computation times by 75%
    • Analyzed extensive geospatial satellite data on high-performance computing systems (Cheyenne and Capser), uncovering global patterns and trends
    • Facilitated development of streamlined sea-surface temperature models through parameter reduction techniques
  • Data Scientist, Lumen Technologies (formerly CenturyLink), June 2020 - August 2020
    • Contributed to the Artificial Intelligence and Machine Learning Center of Excellence during a summer internship
    • Developed and implemented advanced algorithms using timeseries LSTMs and CNNs to predict IT application health and detect anomalies, resulting in improved efficiency and cost savings
    • Automated IT analytics exploratory data analysis (EDA), uncovering previously unknown application issues and driving proactive troubleshooting efforts
    • Generated significant monetary savings through effective identification and resolution of IT application issues
  • Research and Teaching Assistant, Colorado School of Mines, 2018-2020
    • Developed and released practical and theoretical tools to analyze highly nonstationary environmental data sets on the order of hundreds-of-millions of observations
    • A primary component of my PhD work was as a Research Assistant at the National Center for Atmospheric Research in Boulder, Colorado
  • AmeriCorps Math Fellow, Denver Public Schools, August 2016 - June 2017
    • Enhanced middle school math education by providing personalized and small-group instruction to accelerate students’ learning progress
    • Orchestrated after-school STEM programs, effectively managing resources and activities to promote student engagement and achievement
  • Research Fellow, Four-College Biomathematics Consortium, May 2015 - May 2016
    • Conducted innovative research on bovine water-intake in pasture, resulting in valuable insights for the industry
    • Developed cost-effective and efficient data collection tools using Arduino and C++, enhancing data accuracy and streamlining processes
    • Utilized advanced Matlab modeling techniques to analyze ice phenology of Lake Linné, Svalbard, contributing to the understanding of Arctic ecosystems
    • Actively shared research findings at consortium conferences, fostering collaboration and knowledge exchange within the scientific community

Computer Skills

  • Programming: Python, R, Matlab, C++, Bash/Shell script, LaTeX, PostgreSQL
  • Packages: Numpy, Scipy, Pandas, scikit-learn, Keras, Tensorflow, ggplot2, dplyr, cartopy, netCDF4, dask, xarray, iris, geopandas, pytest
  • Other: Git, MPI Programming, CI/CD

Software Publications and Contributions

  • Matlab: DeepTreeMRA, MRA-Parallel, MRA-Serial (Implementations of the Multi-resolution Approximation spatial model for various computational infrastructures)

  • Python: pyaerocom (Python tools for the AeroCom project), optimparallel (Parallel computing interface to the L-BFGS-B optimizer)

Professional Services

  • Referee: Electronic Journal of Statistics, The Annals of Applied Statistics, Environmetrics
  • Judge: Colorado School of Mines Undergraduate Research Symposium 2020

Peer-Reviewed Publications

Technical Reports

Poster Publications

Manuscripts in Preparation and Preprints

Conference Contributions & Talks

Teaching